Quantitative Analysis of Acrylamide Labeled Serum Proteins by LC

Quantitative Analysis of Acrylamide Labeled Serum Proteins by LC-MS/MS Vitor Faca,* Marc Coram, Doug Phanstiel, Veronika Glukhova, Qing Zhang, Matthew Fitzgibbon, Martin McIntosh, and Samir Hanash Fred Hutchinson Cancer Research Center, 1100 Fairview Ave. N., Seattle, WA - 98109 Received March 17, 2006

Abstract: Isotopic labeling of cysteine residues with acrylamide was previously utilized for relative quantitation of proteins by MALDI-TOF. Here, we explored and compared the application of deuterated and 13C isotopes of acrylamide for quantitative proteomic analysis using LC-MS/ MS and high-resolution FTICR mass spectrometry. The method was applied to human serum samples that were immunodepleted of abundant proteins. Our results show reliable quantitation of proteins across an abundance range that spans 5 orders of magnitude based on ion intensities and known protein concentration in plasma. The use of 13C isotope of acrylamide had a slightly greater advantage relative to deuterated acrylamide, because of shifts in elution of deuterated acrylamide relative to its corresponding nondeuterated compound by reversedphase chromatography. Overall, the use of acrylamide for differentially labeling intact proteins in complex mixtures, in combination with LC-MS/MS provides a robust method for quantitative analysis of complex proteomes. acrylamide isotope labeling • LTQ-FTICR • human serum

Introduction Numerous methods have been introduced for quantitative analysis of proteins by mass spectrometry (MS). The choice of method for MS-based quantitation depends on the nature of the application and type of sample of interest. Intact cells can be labeled in vivo using cell culture media enriched with 15N, or stable isotopes of amino acids. Alternatively, labeling can be performed with various reagents during or after enzymatic digestion, although samples to be compared have to be processed separately until the labeling step, which may introduce artifactual variations. Quantitative proteomics is reviewed in Ong and Mann, Sechi and Oda, and Julka and Regnier.1-3 Plasma proteome analysis represents a major challenge for quantitative proteomics because of the wide range of protein concentration and the occurrence of multiple isoforms that may need to be separately quantified. Our group has implemented an intact-protein based approach to serum profiling, with extensive fractionation of nondigested proteins to reduce sample complexity prior to mass spectrometry and to allow separation of isoforms and increased depth of analysis.4 Among * To whom correspondence should be addressed. Tel: (206) 667-1943. Fax: (206) 667-2537. E-mail: [email protected]. 10.1021/pr060102+ CCC: $33.50

 2006 American Chemical Society

the amino acids that could be tagged in an intact-protein based approach, cysteine is a good target because it occurs in some 96% of all human proteins and in ∼27% of tryptic peptides.5 Cysteine is efficiently alkylated by several classes of reagents,6 and has been widely used in protein chemistry to facilitate enzymatic digestion and to prevent protein refolding.9 Because of these advantages, the cysteine alkylation reaction has been extensively used to tag proteins with stable isotopes for mass spectrometry based quantitative analysis.7-10 While cysteine alkylation with acrylamide is an undesired reaction that frequently occurs during polyacrylamide gel electrophoresis,11 there are several properties that make acrylamide a very useful tagging reagent for quantitative studies that rely on isotope labeling. First, it is a small reagent (mass ) 71) that does not introduce significant mass shift or charge changes in the protein and does not negatively affect protein solubility, since it is an hydrophilic tag; second, the mass shift at the peptide level is also minimal, resulting in relative simple MS/ MS spectra, compared to the effect of large tagging reagents such as isotope-coded affinity tags (ICAT);7 third, the reaction is performed using standard protein solubilization solutions and with a virtually 100% yield; additionally the reagents are relatively inexpensive, making it practical to perform experiments starting with large amounts of protein as needed for extensive fractionation and in-depth analysis. Acrylamide alkylation has been previously described as one way to obtain quantitation information by matrix assisted laser desorption/ionization- time-of-flight (MALDI-TOF) for proteins separated by gel electrophoresis.8,9 In this study, we investigated differential labeling of cysteine residues of human serum samples depleted of abundant proteins with acrylamide and two different isotopes (2,3,3′-D3-acrylamide, 1,2,3-13C3-acrylamide), for identification and quantitative analysis using reversed-phase liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) and high-resolution Fourier transform ion cyclotron resonance (FTICR) mass spectrometry. We demonstrated that plasma proteins, most of which are globular and contain several cysteine residues, are quite suitable for alkylation with isotopes of acrylamide. Alkylation with acrylamide did not introduce detectable changes in protein solubility or difficulties in obtaining good collision induced dissociation (CID) spectra of labeled peptides. The use of a high sensitivity and high-resolution mass spectrometer such as the LTQ-FTICR provided excellent MS data, facilitating extraction of reliable relative quantitation for a wide range of serum proteins. Journal of Proteome Research 2006, 5, 2009-2018

2009

Published on Web 07/13/2006

Analysis of Acrylamide Labeled Serum Proteins

technical notes

Materials and Methods Sample Preparation. A pool of sera from three healthy subjects was chromatographically immunodepleted of the top six most abundant proteins using HU-6 columns (4.6 × 100 mm; Agilent, Wilmington, DE) as previously described.6 A total of 400 µg of immunodepleted sample, equivalent to 65µL of the sera pool was concentrated using an Amicon YM-3 device and rediluted in 8 M urea, 30 mM Tris pH 8.5, 0.5% OG (octylbeta-D-glucopyranoside)(w/v). The reduction was performed by adding 0.66 mg dithiotreitol (DTT) per mg of protein and the reaction was carried out at room temperature for 2 h. Samples were split into three aliquots and alkylated with acrylamide, 2,3,3′-D3-acrylamide, 1,2,3-13C3-acrylamide, designated here D0, D3, or 13C-acrylamide. Acrylamide (>99.5% purity) was purchased from Fluka. Isotopes of acrylamide were acquired from Cambridge isotope laboratories (Andover, MA) with minimum chemical purity of 98%. Alkylation was performed by adding 7.1 mg of D0-acrylamide or 7.4 mg of D3 or 13 C-acrylamide per mg of total protein. This protocol was adapted from Sechi et al.8,9 and the amount of acrylamide represents 25-fold molar excess over DTT. The reaction mixture was incubated in the dark for 1 h at room temperature. D3 or 13 C-acrylamide labeled samples were mixed at 1:1 ratio with D0-acrylamide and immediately cleaned-up on a reversedphase trap column (2 × 10 mm packed with Poros R1, Applied Biosystems). The same conditions and reagent proportions were used to reduce and alkylate 50 µg of bovine serum albumin. Different ratios (1:1, 1:3 1:5 and 1:10) of D0 or D3acrylamide alkylated albumin were used to evaluate the method. Protein Digestion And Mass Spectrometric Analysis. Samples were resuspended in 0.25 M urea containing 50 mM ammonium bicarbonate and 4% of acetonitrile (v/v) and digested overnight with 2 µg of modified trypsin (Promega). The digestion was interrupted by addition of 5 µL of 10% formic acid solution (v/v). Samples were analyzed in a LTQ-FTICR mass spectrometer (Thermo-Finnigan) coupled with a nanoAcquity UPLC chromatography system (Waters). Liquid chromatography separation was performed in a 25 cm column (Picofrit 75 µm ID, New Objectives, in house-packed with MagicC18AQ resin) using a 140 min linear gradient from 5 to 40% of acetonitrile in 0.1% formic acid at 250 nL/min or 60 min gradient for bovine serum albumin (BSA). The spectra were acquired in a data-dependent mode in m/z range of 400 to 1800, with selection of the 5 most abundant +2 or +3 ions of each MS spectrum for MS/MS analysis. Mass spectrometer parameters were as follows: capillary voltage of 3.2 KV, capillary temperature of 200 °C, resolution of 100 000 and FT target value of 2 × 106. Acquired data was automatically processed by the Computational Proteomics Analysis SystemsCPAS,12 using the Comet search algorithm. Minimum criteria for peptide matching was Peptide Prophet Score greater than 0.2. Peptides that met these criteria were further grouped to protein sequences using the Protein Prophet algorithm at an error rate of 5% or less.13 For all databank searches, D0-acrylamide alkylation was considered as a fixed modification and heavy isotope labeled peptides were detected using a delta mass of 3.01884 and 3.01006 Da for D3 and 13C-acrylamide, respectively. Quantitation Algorithm. Acrylamide ratios were determined using a script designated “Q3” developed in-house to obtain the relative quantities for each pair of peptides identified by MS/MS that contained cysteine residues. Essentially, the algorithm reconstructs the ion chromatogram for both light 2010

Journal of Proteome Research • Vol. 5, No. 8, 2006

Figure 1. Quantitation of different ratios of BSA. The graph shows the linear distribution of an average of 28 data points for each of the 7 different ratios of BSA analyzed. Although the data indicates greater variability for extreme ratios (1:5, 1:10, and 10: 1) all the data points present good correlation coefficients (R2).

(D0) and heavy (D3 or 13C) forms for each identified peptide containing cysteine residues and computes the intensities for each form. More specifically, from the MS/MS peptide identification list, peptides containing cysteine with Peptide Prophet scores greater than 0.75 were selected. We then obtained the theoretical mass-charge of the monoisotopic light and monoisotopic heavy labeled peptides as well as the theoretical masscharge of each respective 13C isotope peak. Since the mass difference between light and heavy is approximately three times (3.01884 or 3.01006 for D3-acrylamide or 13C-acrylamide, respectively) the number of cysteines present in the peptide, the analysis was restricted to either the monoisotopic and the first 2 13C isotopic peaks of the light form (if there is one cysteine) or the first 5 13C isotopic peaks (if there were two or more cysteines) and their corresponding peaks from the heavy form. The intensity of peaks in each MS scan was computed by centering the MS scan to the location of its maximum closest to the theoretical m/z for each peak within a 25 ppm window. Each light peak was paired with the corresponding heavy peak and if both peaks were present, we considered this a match. In general, since peaks occasionally occur stochastically, only matched peaks contributed to the quantitation. For the final quantitation, we used the primary scan (i.e., the one that immediately preceded the MS/MS identification) as well as a certain number of MS scans immediately before and after the primary scan, up to a limit of 10 scans. For a scan without any matched peaks, this scan and any subsequent scan in both directions was excluded. As a special case applied to low intensity peaks, if no more than one matched pair of ions was found in the primary scan, then all of the isotopic peaks found were used, whether matched or not (i.e., if the monoisotopic peak of both heavy and light form matched, but the heavy form also presented the first 13C isotope, then we considered the sum of both peaks from the heavy and only the monoisotopic peak from the light form in the final quantiation). The resulting ratio was adjusted slightly for all the peptides with mass greater than 1800 Da by subtracting from the derived quantity of the heavy form, the fraction that can be expected to result from overlapping isotopes of the light form. Proteins containing more than one pair of peptides that yielded quantitation had all the ratios

technical notes

Faca et al.

Table 1. Serum Proteins Identified and Quantified using Isotopic Alkylation with Acrylamideg

IPIa

description

1 2 3 4 5 6 7 8 9 10 11 12 13 14

IPI00164623 IPI00478003 IPI00556148 IPI00418163 IPI00555812 IPI00298828 IPI00019580 IPI00017601 IPI00019568 IPI00021841 IPI00022488 IPI00032328 IPI00019591 IPI00021727

15

IPI00294193

16

IPI00022418

17 18 19 20 21 22 23 24 25

IPI00022463 IPI00022895 IPI00022229 IPI00019943 IPI00026314 IPI00022431 IPI00032179 IPI00022434 IPI00305461

26

IPI00292530

27 28 29 30 31 32 33 34 35 36

IPI00304273 IPI00022426 IPI00478493 IPI00550991 IPI00021854 IPI00032291 IPI00022371 IPI00291867 IPI00292950 IPI00479867

37 38 39 40 41

IPI00298971 IPI00009920 IPI00008558 IPI00022432 IPI00007240

42 43 44

IPI00022429 IPI00020091 IPI00011252

45 46

IPI00291262 IPI00291866

47 48

IPI00166729 IPI00021885

49

IPI00017696

50 51 52 53

IPI00025426 IPI00296608 IPI00022395 IPI00163207

54 55 56 57 58


complement C3 precursor alpha-2-macroglobulin precursor complement factor H complement C4 precursor vitamin D-binding protein, beta-2-glycoprotein I precursor plasminogen precursor ceruloplasmin precursor prothrombin precursor apolipoprotein A-I precursor hemopexin precursor kininogen precursor complement factor B precursor C4b-binding protein alpha chain precursor inter-alpha-trypsin inhibitor heavy chain H4 precursor fibronectin; cold-insoluble globulin; migration-stimulating factor serotransferrin precursor* alpha-1B-glycoprotein precursor apolipoprotein B-100 precursor afamin precursor gelsolin precursor alpha-2-HS-glycoprotein precursor antithrombin-III precursor serum albumin precursor inter-alpha-trypsin inhibitor heavy chain H2 precursor inter-alpha-trypsin inhibitor heavy chain H1 precursor apolipoprotein A-IV precursor AMBP protein precursor haptoglobin precursor e alpha-1-antichymotrypsin precursor apolipoprotein A-II precursor complement C5 precursor histidine-rich glycoprotein precursor complement factor I precursor heparin cofactor II precursor similar to Complement C1r component precursor vitronectin precursor complement component C6 precursor plasma kallikrein precursor transthyretin precursor coagulation factor XIII B chain precursor alpha-1-acid glycoprotein 1 precursor alpha-1-acid glycoprotein 2 precursor complement component C8 alpha chain precursor clusterin precursor plasma protease C1 inhibitor precursor alpha-2-glycoprotein 1, zinc fibrinogen alpha/alpha-E chain precursor complement C 1s subcomponent precursor pregnancy zone protein precursor complement component C7 precursor complement component C9 precursor N-acetylmuramoyl-L-alanine amidase precursor apolipoprotein E precursor platelet basic protein precursor properdin precursor apolipoprotein D precursor complement component C8 beta chain precursor

gene name

C3 A2M CFH C4 GC APOH PLG CP F2 APOA1 HPX KNG BF C4BPA

concentration in plasma (ng/mL)b

9.5E+05 1.4E+06 1.7E+05 1.4E+05 2.1E+05 1.4E+06 7.5E+05

D3/D0

13C/D0

unique peptidec

D3/D0 ratiod

ratio stdevd

unique peptidec

13C/D0 ratiod

ratio stdevd

175 126 102 114 88 53 76 81 63 47 55 51 48 44

1.17 1.14 1.13

0.12 0.08 0.06

1.02

0.01

0.96

0.16

1.17 1.19 1.14 1.13

0.01 0.08 0.00 0.03

1.06 1.02 1.08 1.00

0.04 0.04 0.05 0.04

1.08 1.12 1.18 1.38

0.02 0.05 0.10 0.18

90 89 110 80 87 79 56 43 54 70 47 48 49 39

1.02 1.05 1.04 0.93

0.03 0.11 0.03

ITIH4

51

FN1

45

0.89

0.13

35

1.12

0.18

49 29 63 35 29 29 37 43 36

1.25 1.15 1.04 1.11 1.25 1.07 0.92 1.05 0.76

0.11 0.13 0.03 0.16 0.12 0.02 0.09 0.03

30 45 10 27 31 30 19 12 15

1.03 1.01 0.94 1.12 0.89 1.04 0.95 0.96 0.77

0.04 0.03 0.02 0.24 0.06 0.03 0.08 0.19

ITIH1

38

0.97

0.03

12

APOA4 AMBP HP APOA2 C5 HRG IF SERPIND1 C1R

24 28 22 37 19 37 27 26 21 29

1.18 1.40

0.11 0.46

1.17 0.95 1.00 0.97 1.22 1.07

0.01 0.08 0.08 0.02 0.43 0.04

21 23 22 16 25

1.10 1.29 1.11

0.07 0.24 0.02

1.07

17 18 21

CLU SERPING1

TF A1BG APOB AFM GSN AHSG SERPINC1 ALB ITIH2

2.3E+06 7.2E+05

3.2E+05 4.0E+07

31

25 21 24 8 27 7 17 17

0.95 1.08

0.01

1.20 1.00 0.98 0.88

0.08 0.03 0.14

13

1.01

0.16

0.93 1.12 1.01

0.03 0.19 0.04

0.06

20 18 16 22 12

1.00

0.07

1.00 1.07 0.90

0.10 0.02 0.03

19 17 14

0.98 0.94 1.15

0.02 0.04 0.43

15 21

1.41 1.21

0.30 0.20

17 12

1.03

0.01

AZGP1 FGA

14 17

1.02 1.20

0.06 0.24

17 12

0.95 0.94

0.00

C1S

15

1.26

0.03

14

1.06

0.03

PZP C7 C9 PGLYRP2

13 16 15 14

1.07 1.08 1.05 1.03

0.17 0.03 0.19 0.01

9 10 11

1.05 1.06 1.01

0.27 0.20 0.12

16 10 6 8 13

1.05 1.05 0.95 1.08

0.06 0.02 0.05 0.16

8 13 17 15 9

0.79 1.07 0.96 0.88 0.92

0.08 0.05 0.07 0.10 0.04

VTN C6 KLKB1 TTR F13B ORM1 ORM2 C8A

APOE PPBP PFC APOD C8B

8.8E+05 3.0E+05

2.6E+05 6.1E+05 6.1E+05

3.4E+04

Journal of Proteome Research • Vol. 5, No. 8, 2006 2011

technical notes

Analysis of Acrylamide Labeled Serum Proteins Table 1. (Continued)

D3/D0

59 60 61 62 63 64 65 66 67 68 69 70

IPIa

description

IPI00009028 IPI00022420 IPI00385058 IPI00218732 IPI00294004 IPI00032220 IPI00019581 IPI00027235 IPI00218816 IPI00022446 IPI00029863 IPI00011261

tetranectin precursor plasma retinol-binding protein precursor hypothetical protein serum paraoxonase/arylesterase 1 vitamin K-dependent protein S precursor angiotensinogen precursor coagulation factor XII precursor attractin precursor HBB protein Row70Platelet factor 4 precursor alpha-2-antiplasmin precursor complement component C8 gamma chain precursor complement component 1, q subcomponent, insulin-like growth factor binding protein complex acid labile chain MGC27165 protein serum amyloid A-4 protein precursor apolipoprotein M complement C2 precursor complement C1q subcomponent, A chain precursor pigment epithelium-derived factor precursor thyroxine-binding globulin precursor kallistatin precursor apolipoprotein C-II precursor carboxypeptidase N 83 kDa chain (regulatory subunit) lumican precursor apolipoprotein C-III precursor hemoglobin alpha-1 globin chain leucine-rich alpha-2-glycoprotein precursor apolipoprotein(a) precursor fibulin-1 precursor apolipoprotein C-I precursor cholinesterase precursor complement factor H-related protein 3 precursor HGF activator like protein Ig gamma-2 chain C region IGHM protein C4b-binding protein beta chain precursor FLJ00385 protein hypothetical protein keratin, type II cytoskeletal 1 fetuin-B precursor haptoglobin-related protein hepatocyte growth factor activator precursor alpha-1-antitrypsin precursor biotinidase precursor [Homo sapiens] extracellular matrix protein 1 precursor megakaryocyte stimulating factor selenoprotein P precursor corticosteroid-binding globulin precursor IGLC1 protein keratin, type I cytoskeletal 10 phosphatidylinositol-glycan-specific phospholipase D 1 precursor apolipoprotein-L1 precursor complement factor H-related protein 5 precursor hepatocyte growth factor-like protein precursor insulin-like growth factor binding protein 3 precursor

71 IPI00218746 72 IPI00020996 73 74 75 76 77


78 IPI00006114 79 80 81 82

IPI00292946 IPI00328609 IPI00021856 IPI00166930

83 84 85 86

IPI00020986 IPI00021857 IPI00410714 IPI00022417

87 88 89 90 91


92 93 94 95

IPI00041065 IPI00399007 IPI00382937 IPI00025862

96 97 98 99 100 101

IPI00168728 IPI00386158 IPI00220327 IPI00005439 IPI00296170 IPI00029193

102 IPI00305457 103 IPI00218413 104 IPI00003351 105 IPI00024825 106 IPI00029061 107 IPI00027482 108 IPI00154742 109 IPI00009865 110 IPI00299503 111 IPI00177869 112 IPI00006543 113 IPI00292218 114 IPI00018305 2012


gene name

13C/D0

concentration unique D3/D0 ratio unique 13C/D0 ratio in plasma peptidec ratiod stdevd peptidec ratiod stdevd (ng/mL)b

TNA RBP4 PON1 PROS1 AGT F12 ATRN HBB PF4 SERPINF2 C8G

16 14 11 14 14 15 12 13 11 7 12 13

1.10 1.13 1.05 1.33 1.01

0.04 0.04 0.04 0.11 0.06

1.06 0.98 1.11 1.06 0.95 1.00

0.01 0.16 0.11 0.04 0.08

6 6 9 6 6 4 7 5 8 11 5 4

1.08 0.90 1.07 0.95 1.23

0.07 0.02 0.03 0.03 0.28

1.18 0.85

0.15 0.03

1.03 0.84 1.00

0.03 0.01

-

12

1.06

0.06

4

IGFALS

13

1.43

0.34

3

1.02

0.13

IGHA1 SAA4 APOM C2 C1QA

9 6 9 12 7

1.09

0.08

7 10 6 3

1.05

0.07

1.22 1.25 1.10

0.00

SERPINF1

10

1.13

SERPINA7 SERPINA4 APOC2 -

12 12 7 11

LUM APOC3 HBA1 LRG1

11 6 6 7

LPA FBLN1 APOC1 BCHE CFHL3

5 7 6

1.15 1.03

4

1.01

0.01

0.93

0.20

2 2 6 2

0.98

0.11

2 7 7 5

0.99 1.07

0.02

3

1.00 1.10

HABP2 IGHG2 C4BPB

6 6 6 6

FLJ00385 KRT1 FETUB hpr HGFAC

1.51 0.89 0.75

0.04

0.03 0.03

6 4 5 3 8

1.01

0.03

1.01 1.02 0.95 1.03

0.09 0.05 0.02 0.10

5 5 5 4

0.92 0.96 1.23 0.87

0.03 0.06 0.34 0.13

7

1.08

0.02

4 5

0.97

0.01

5 7 9 7

1.37 1.21 1.04

0.09 0.04 0.04

3 1 3

1.26 1.01 0.91

0.03 0.09 0.10

SERPINA1 BTD ECM1

8 5 5

0.97 1.15

4

1.45

0.16

PRG4 SEPP1 SERPINA6

3 4 4

6 5

0.83

IGLC1 KRT10 GPLD1

5 4 6

4

1.08

0.05

APOL1 CFHL5

5 5

1.39

3 3

1.01

0.08

MST1

5

1.23

3

0.89

IGFBP3

8

5.9E+01

3

2 0.11

3 0.26

5

technical notes

Faca et al.

Table 1. (Continued)

IPIa

description

115 IPI00296176 coagulation factor IX precursor 116 IPI00006154 complement factor H-related protein2 precursor 117 IPI00061246 hypothetical protein 118 IPI00026199 plasma glutathione peroxidase precursor 119 IPI00299435 apolipoprotein F [Homo sapiens] 120 IPI00022937 coagulation factor V 121 IPI00008556 coagulation factor XI precursor 122 IPI00297550 coagulation factor XIII A chain precursor 123 IPI00293925 ficolin 3 precursor 124 IPI00028413 inter-alpha-trypsin inhibitor heavy chain H3 precursor 125 IPI00032311 lipopolysaccharide-binding protein precursor 126 IPI00027350 peroxiredoxin 2 127 IPI00007221 plasma serine protease inhibitor precursor 128 IPI00419630 tax_Id ) 9606 DPKL1915 129 IPI00296099 thrombospondin-1 precursor 130 IPI00021817 vitamin K-dependent protein C precursor 131 IPI00022394 complement C1q subcomponent, C chain precursor 132 IPI00022391 serum amyloid P-component precursor 133 IPI00028030 cartilage oligomeric matrix protein precursor 134 IPI00011264 complement factor H-related protein 1 precursor 135 IPI00004798 cysteine-rich secretory protein-3 precursor 136 IPI00019359 keratin, type I cytoskeletal 9 137 IPI00218795 L-selectin precursor 138 IPI00022368 Row70Serum amyloid A protein precursor 139 IPI00023019 sex hormone-binding globulin precursor 140 IPI00550315 Ig kappa chain C region 141 IPI00001611 insulin-like growth factor II precursor 142 IPI00022731 apolipoprotein C-IV precursor 143 IPI00004656 beta-2-microglobulin precursor 144 IPI00027462 calgranulin B 145 IPI00329775 carboxypeptidase B2 precursor 146 IPI00025204 CD5 antigen-like precursor 147 IPI00019576 Coagulation factor X precursor 148 IPI00242956 IgG Fc binding protein [Homo sapiens] 149 IPI00247295 nesprin 1 150 IPI00022331 phosphatidylcholine-sterol acyltransferase precursor 151 IPI00007199 protein Z-dependent protease inhibitor precursor 152 IPI00029699 Ribonuclease 4 precursor 153 IPI00178926 similar to immunoglobulin J chain 154 IPI00009793 complement C1r-like proteinase 155 IPI00023673 galectin-3 binding protein precursor 156 IPI00550640 IGHG4 protein 157 IPI00019038 lysozyme C precursor 158 IPI00029260 monocyte differentiation antigen CD14 precursor 159 IPI00003590 quiescin Q6 160 IPI00395488 vasorin 161 IPI00374068 thrombospondin repeat containing 1 162 IPI00020019 adiponectin precursor 163 IPI00452748 amyloid protein A 164 IPI00008554 angiogenin precursor 165 IPI00215983 carbonic anhydrase I

gene name

concentration unique in plasma peptidec (ng/mL)b

D3/D0 D3/D0 ratiod

13C/D0 ratio unique stdevd peptidec

13C/D0 ratiod

ratio stdevd

F9 CFHL2

5 4

1.13 1.06

0.07 0.03

2 4

0.92 1.04

0.01

GPX3

5 4

1.01 0.97

0.02 0.15

2

1.04

0.06

APOF F5 F11 F13A1

3

1.13

0.15

3 4

0.81 0.96

3 3

1.09

0.02

FCN3 ITIH3

3 5

1.00

LBP

3

PRDX2 SERPINA5

3 3

THBS1 PROC

3 4 3

C1QG

4

2

APCS

4

2

COMP

3

1.09

0.17

CFHL1

2

0.90

0.01

3

0.96

0.13

CRISP3

3

1.00

0.04

3 3

1.01

0.04

1 3

0.73

2

0.91

KRT9 L-selectin SAA2

1.7E+01

SHBG

4

IGKC IGF2

2 3

APOC4 B2M S100A9 CPB2 CD5L F10 -

1.1E+03

3 2 2 2 2 2 2

SYNE1 LCAT

2 2

SERPINA10

2

RNASE4 C1RL LGALS3BP

2 3 2 2

LYZ CD14

2 2 2

QSCN6 MRX85 TSRC1

2 2

ADIPOQ SAA1 ANG CA1

1 1

0.02

2

2 1.07

0.13

0.99 0.97

0.69

3 2

1.09 0.94

0.09 0.09

1.04

0.03

2 2

1.05

0.07

1 1 2

1.06 0.07 5.51(0.93)f 0.76

2 2 1

1.16

1 1

0.91

1.15

0.03

4.92 (0.78)f 0.95 0.06

1.01

2 1 1

1 1 Journal of Proteome Research • Vol. 5, No. 8, 2006 2013

technical notes

Analysis of Acrylamide Labeled Serum Proteins Table 1. (Continued)

IPIa

description

166

IPI00010295

167

IPI00029658

168

IPI00218834

169 170 171 172

IPI00473011 IPI00301143 IPI00385985 IPI00009477

173

IPI00216651

174

IPI00384401

175 176

IPI00299059 IPI00022733

177

IPI00013179

178

IPI00216882

179

IPI00018136

carboxypeptidase N catalytic chain precursor EGF-containing fibulin-like extracellular matrix protein 1 precursor Fc of IgG, low affinity IIIa, receptor for hemoglobin delta chain hypothetical protein PSEC0164 Ig lambda chain V-III region LOI intercellular adhesion molecule-2 precursor interleukin-28 receptor alpha chain precursor myosin-reactive immunoglobulin kappa chain variable region neural cell adhesion molecule phospholipid transfer protein precursor prostaglandin-H2 D-isomerase precursor similar to mannan-binding lectin serine protease 1 vascular cell adhesion protein 1 precursor

gene name

CPN1

concentration in plasma (ng/mL)b

D3/D0 unique peptidec

13C/D0 ratio stdevd

unique peptidec

13C/D0 ratiod

1

0.90

ratio stdevd

1

EFEMP1 FCGR3A

1

HBD PI16 ICAM2

1 1 1

IL28RA

1

1 1

-

1

1

CHL1 PLTP

1 1

PTGDS

1

-

1

VCAM1

D3/D0 ratiod

1

a The databank IPI version 3.09 was used for protein identification. b Values reproduced from Haab et al.14 c Different charge states and different states of oxidation as well as heavy or light form of the peptide were considered different peptides in deriving the final number of unique peptides. d Standard deviation of ratios was calculated for the protein ratios across duplicates. e Residual protein from immunodepletion. f Value generated automatically (value manually calculated). g A total of 160 proteins were identified with more than 2 peptides in all 4 replicates of unfractionated serum; 113 proteins (70%) presented at least one peptide containing cysteine which provided quantitation information.

averaged. This quantitation algorithm will be incorporated in the near future to the open source distributions of the Mass Spectrometry in silico Peptide Characterization Tool (msInspect)15 (http://proteomics.fhcrc.org) and integrated into the Computational Proteomics Analysis System (CPAS).12

Results and Discussion Alkylation Of Cysteines With Acrylamide. The feasibility of acrylamide labeling for protein identification and quantitation has been demonstrated previously using MALDI-TOF for proteins separated by two-dimensional electrophoresis,10,11 but not with high-throughput LC-MS/MS. In the present work, alkylation with acrylamide, 2,3,3′-D3-acrylamide, 1,2,3-13C3acrylamide, designated here D0, D3, or 13C-acrylamide, respectively, was investigated with bovine serum albumin and applied to human immunodepleted serum. For both types of samples, no protein precipitation was observed, indicating that acrylamide based alkylation is compatible with intact-protein based approaches. A very high yield of cysteine alkylation was achieved, since databank searches for nonmodified cysteines did not return matches and we could not observe nonalkylated peptides by manual inspection of spectra. However, for several ion pairs of D0/D3 or D0/13C-acrylamide labeled peptides, we observed a small set of adducts also in pairs separated by 3 Da, corresponding to less than 5% of the main pair. These adducts have mass increments of +16, +32, and +48. Analysis of the MS/MS spectra of these adducts (data not shown) indicated that addition of +16 and +32 occurred in the Cyspropionamide residue. Our explanation for these adducts would be oxidation of the labeled cysteine residues as observed for methionine during the electrospray ionization, forming Cys2014


propionamide sulfoxide (+16) or Cys-propionamide sulfone (+32). However, the third state of oxidation occurs at another site, since the sulfur atom of the Cys-propionamide could not accommodate one more atom of oxygen. Bovine serum albumin is a protein rich in cysteine residues and represents a good standard to evaluate the approach presented here. Seven different D0/D3 labeled albumin ratios were evaluated. An average of 28 cysteine residues, out of 35 present in BSA, provided quantitation information for the ratios tested. Exemples of spectra obtained for 7 different ratios of D0/D3 acrylamide are shown in supplementary data. The result was obtained in a LTQ-FTICR mass spectrometer at 100 000 resolution, as utilized for analysis of complex samples in the data dependent acquisition mode. The spectra were welldefined and showed clearly the presence of both envelopes of ions for D0 and D3 labeled peptides even for ratios 10:1 (supplementary figure). Such resolution is important to resolve individual isotopic peaks from the D0 or D3-labeled peptide, directly impacting on peptide identification and accuracy of quantitation and on separation of different coeluting peptides especially in high complex mixtures such as human biological samples. Also the high-resolution ensures that background noise has minimal interference in quantitation. Figure 1 shows the plot of all quantitation data points obtained, illustrating the linearity of this approach. The coefficient of variation calculated for all the albumin peptide ratios for the D0/D3 pairs 1:3, 1:1, and 3:1 were respectively 17.3%, 13.2%, and 19.2%, indicating accurate quantitation. These values are similar to those obtained in a recent ICAT evaluation that resulted in a coefficient of variation of 18.6%.16 For ratios above 10-fold, these variations increased significantly to about

technical notes 35-50%, especially for ratios where the peptide labeled with the light form of acrylamide is more intense, due to interference of isotopic peaks representing 3 or 4 13C. Similar results have been reported for other quantitation methods in which the separation of light and heavy envelopes is only 3 Da.17,18 Although this decrease in quantitative accuracy for extreme ratios (.10-fold) is significant, the method is still valuable for assessing large biological quantitative variations. For these particular cases, occasionally only the light or heavy form of the peptide is detected (unpublished datasnot shown), which have to be individually evaluated to avoid distortions in the final calculation of the protein ratio. Analysis Of Immunodepleted Human Serum. To assess the merits of isotopic acrylamide labeling of intact proteins, a pool of serum from three healthy subjects was immunodepleted to remove the top six most abundant proteins, after which the remainder of the proteins were reduced with DTT, split into three aliquots and labeled with D0, D3-acrylamide, or 13Cacrylamide. Mixtures of equal amounts of D0 and D3 or D0 and 13C-labeled aliquots were analyzed in duplicate using a 2 h reversed-phase gradient in a LTQ-FTICR mass spectrometer. Analysis of immunodepleted but otherwise unfractionated serum labeled with D0/D3 or D0/13C yielded 160 proteins identified with two or more peptides in replicate runs (Table 1). A total of 121 of these proteins (76%) were present in four replicates, indicating good reproducibility of the overall method. Among the proteins identified in this study, complement C3 is found in plasma at 1 mg/mL, β-2-microglobulin at 1 µg/mL and insulin-like growth factor binding protein 3 precursor at 60 ng/mL,14 which corresponds to more than 5 orders of magnitude in protein concentration in a single run. This is particularly noteworthy given that no fractionation step was applied.4 Quantitation ratios were calculated for those peptides containing cysteine residues identified by MS/MS, with a Peptide Prophet score >0.75. The presence of a coeluting pair of ions separated by multiples of 3 Da also indicates correct assignment of the peptide sequence containing cysteine.8 Cysteine containing peptides provided quantitation information for a total of 113 proteins (69 present on all four replicates) (Table 1). Several proteins presented multiple peptides containing cysteine as observed in Figure 2 for plasminogen, for which out of 35 peptides containing cysteine in the sequence, 24 were detected in this study. Among the 2493 unique peptides identified (unique sequences) in our entire data set (duplicates of D0/D3 and D0/13C), 857 peptides (34.4%) contained at least one cysteine residue. This value is slightly higher than the estimate of tryptic peptides containing cysteine predicted based on the human genome (26.6%).5 It is likely that serum proteins are particularly rich in cysteine residues compared for example to membrane proteins. Quantitation of individual proteins, represented by a particular IPI number, was calculated by averaging the ratios obtained for cysteine labeled peptides. In particular cases for which one peptide ratio represented an outlier among several others from the same protein, this ratio was excluded from the final protein ratio computation. As can be observed in Table 1, few replicates presented standard deviations for duplicates greater than 0.25. Only two proteins, coagulation factor X precursor and Ribonuclease 4, each containing a single peptide ratio measurement, presented unexpected high values. Manual inspection of these outliers indicated that for coagulation factor X precursor and ribonuclease 4, incorrect ratios were computed

Faca et al.

Figure 2. Quantitation of plasminogen. Of the 35 peptides containing cysteine in plasminogen, 24 were detected and provided quantitation data. The standard deviations calculated for the data (at the bottom) indicates that the pair of acrylamide isotopes D0/13C provides highly accurate data. Journal of Proteome Research • Vol. 5, No. 8, 2006 2015


technical notes

Figure 3. Quantitation of serum proteins. A mixture of 1:1 immunodepleted human serum sample was labeled with D0/D3-acrylamide (right panels) or D0/13C-acrylamide (left panels). These data points represent an average of two measurements per peptide and this redundancy in quantification results from redundancy in peptide identification, which increases confidence in the data. Panel A illustrates the distribution of quantitation data correlated to peptide mass. Most of the data points (89%) were acquired for peptides with mass lower than 3000 Da. The plots in panel B show the distribution of ratios obtained for intensities of the heavy and light forms of acrylamide. The points are linearly distributed over 5 orders of magnitude based on ion intensities. Panel C shows the distribution of quantitation ratios obtained for the pairs D0/D3 and D0/13C. 94 and 96% of the data points are within a 2-fold ratio threshold (indicated by dashed lines) and 86 and 94% within a 1.5-fold for D0/D3 and D0/13C, respectively.

due to low intensities of the peptide ions (s/n < 4) and ratios manually calculated from peak intensities were respectively 0.93 and 0.78. Also, it can be observed in Table 1 that often the number of peptides obtained for the pair D0/D3 was greater than for D0/13C. We attribute this difference solely to a lower recovery of the D0/13C proteins during samples cleanup, since no chemical reactivity difference can be expected from acrylamide isotopes and the purity grade of the reagents was similar (>98%). A complicating factor with the acrylamide labeling method is overlapping of light and heavy labeled isotopes, due to the low mass difference (3 Da). However, this effect would be more noticeable for peptides with mass above 3000 Da, for which the relative intensity of the third 13C isotope of the light form 2016


could contribute about 46% of the intensity of the monoisotopic peak of the heavy form. In the present work, the script that calculates intensities of the quantitation pairs took into consideration the predicted contribution of the third, fourth and fifth 13C isotope of the light form which was subtracted from the corresponding ions of the heavy form envelope for all peptides with single cysteine residues and mass greater than 1800. As can be noted in Figure 3a, peptides above 3000 Da provided reliable data after this correction for both D0/D3 and D0/13C-acrylamide pairs. Also it would be expected that the larger the peptide, the greater the chance for occurrence of more than one cysteine residue. In this study, we detected peptides containing up to 4 cysteine residues, although the number of these high mass peptides was small, due to

technical notes

Faca et al.

Figure 4. Influence of deuterated acrylamide peptide elution. The reconstructed ion chromatograms compare the elution profiles of the peptide VCPFAGILENGAVR from β-2-glycoprotein I precursor labeled with the pairs D0/D3 (upper panel) and D0/13C (lower panel). The elution profile of the heavy form (D3 or 13C-acrylamide) is represented in gray and in black for the light form (D0 acrylamide). Early elution of the D3-acrylamide labeled peptide is noticeable interfering with the quantitation ratios obtained at different retention times. Spectra on the right correspond to the D0/D3 or D0/13C pairs obtained at the critical retention times indicated with arrows.

limitations in ionization and fragmentation as well as MS/MS data complexity. As previously indicated, proteins identified in this study occur in human plasma in concentrations ranging from mg/ mL to ng/mL (Table 1). The levels of ion intensities for peptides that provided quantitation information were also distributed across some 5 orders of magnitude as illustrated in Figure 3b. This correspondence indicates that the combination of a high accuracy and sensitivity mass spectrometer and simple and efficient isotopic tagging with acrylamide provides a robust method for quantitative analysis of complex proteomes. Comparison Of D3-Acrylamide And 13C-Acrylamide. A concern with the use of deuterated compounds for quantitation in LC-MS is a small difference in hydrophobicity between deuterated and nondeuterated pairs. This difference may result in small shifts in chromatographic retention time of peptides as observed with the first generation of ICAT reagents in which peptides labeled with the deuterated tag eluted some seconds earlier than the nondeuterated tag that contains only hydrogen.19 This effect is not expected for isotopes containing 13C. To evaluate the consequence of these effects on quantitation results, we compared the performance of D0/D3 and D0/13C pairs using data obtained for serum proteins.

Hydrophobicity of the domain into which heavy isotopes are incorporated is the most important indicator of whether a coding agent will exhibit a deuterium isotope effect during reversed-phase chromatography.19 Deuterium atoms in a polar and small functional group such as acrylamide would be expected to present a small resolution effect. We observed consistent differences in retention time of about 1-2 s between the D3 and the D0-acrylamide across the entire chromatogram. Figure 4 presents the elution profile of one peptide containing a single cysteine residue with about a 6 s-shift between D0/D3 labeled peaks. The quantitation script takes the MS/MS sequencing event as the starting point to extract the ion intensities and a 10-MS scan (∼14 s) window is considered for the calculations. Consequently, significant shifts can lead to errors. For this particular example, a total of 6 ratios (6 MS/MS events for the peptide) were obtained for the peptide as a result of wide elution of the corresponding peak. The values have an average of 1.06 (D3/D0), but when considered individually, we observed influence of the shift in elution into the quantitation (1.85, 1.02, 1.33, 0.87, 0.60, 0.72). Points taken at the beginning of the elution had greater contribution of the D3-labeled isotope, increasing the ratio, while points at the end had greater contribution of the D0-labeled ion (see Figure 3 left panels). This same peptide, labeled with the pair D0/13C, is illustrated Journal of Proteome Research • Vol. 5, No. 8, 2006 2017

technical notes


in Figure 4. As can be observed, no difference in elution of isotopes of acrylamide can be detected for the pair D0/13C as predicted. Several quantitation data points were also obtained for the D0/13C-labeled peptide, due to its broad elution profile. In contrast to D0/D3 findings, basically the same ratio was obtained across all elution points (1.07, 1.01, 0.98, 1.05, 1.05, 0.90, 1.00, 0.92, 0.99) for the D0/13C-labeled peptide. The standard deviation for the ratios was (0.06, compared to (0.46 for the D0/D3 pair. Other undesirable effects resulting from distinct elution of deuterated and nondeuterated peptides may complicate analysis since they can be ionized at different times or may overlap with other peptides with similar mass, increasing the possibility of producing both systematic and random quantitation errors.19 These effects could be reflected in quantitation data obtained across the entire experiment. Figure 2 illustrates peptide quantitation findings for plasminogen. Greater quantitative variability was observed for D0/D3, with a standard deviation that was 2-fold higher than that obtained for D0/13C-acrylamide. We attribute this variability to the deuterium effect, since the samples compared were the same. In the same way, the influence of resolution of deuterated compounds is evident in all the panels presented in Figure 3. Though both D0/D3 and D0/13C pairs presented almost the same modes of 1.13 and 1.11 (Figure 3C) and percentage of data points within a 2-fold threshold (94% and 96%), when this threshold was lowered to 1.5-fold, the differences became significant (86% and 94%). In both the correlation ratio/peptide mass (Figure 3a) and the distribution plot of the quantitation data (Figure 3b), there was evident spreading of values for the pair D0/D3. On the basis of all the comparisons, we can conclude that for more accurate analysis, the D0/13C-acrylamide pair should be used, though the D0/D3 pair can also provide useful data.

Conclusions The quantitative LC-MS/MS approach presented is simple, low cost and takes advantage of the high-sensitivity/high resolution provided by an instrument like the LTQ-FTICR mass spectrometer and provides both high confidence identification and quantitative data for complex protein mixture. Differential alkylation of intact serum proteins with acrylamide isotopes provided relative quantitation for about 70% of proteins identified (113 proteins quantified from 160 proteins identified with two or more peptides). These proteins were distributed over 5 orders of magnitude of abundance levels based on the range of intensity of peptide ions and known protein concentration in plasma. The variations caused by chromatographic resolution of deuterated and nondeuterated acrylamide labeled peptides lead to the conclusion that use of 13C isotope of acrylamide is preferred because of greater accuracy. Implementation of this methodology in combination with intact protein separation methods would be a valuable tool for the analysis of complex biological fluids proteomes.

Acknowledgment. We thank Salvatore Sechi and Jimmy Eng for helpful advice regarding this study. Supporting Information Available: High resolution mass spectra of D0 and D3 labeled peptide. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Ong, S. E.; Mann, M. Mass spectrometry-based proteomics turns quantitative. Nat. Chem. Biol. 2005, 1, 252-262.

2018


(2) Sechi, S.; Oda, Y. Quantitative proteomics using mass spectrometry. Curr. Opin. Chem. Biol. 2003, 7, 70-77. (3) Julka, S.; Regnier, F. Quantification in proteomics through stable isotope coding: a review. J. Proteome Res. 2004, 3, 350-363. (4) Wang, H.; Clouthier, S. G.; Galchev, V.; Misek, D. E.; Duffner, U.; Min, C. K.; Zhao, R.; Tra, J.; Omenn, G. S.; Ferrara, J. L.; Hanash, S. M. Intact-protein-based high-resolution three-dimensional quantitative analysis system for proteome profiling of biological fluids. Mol. Cell. Proteomics 2005, 4, 618-625. (5) Zhang, H.; Yan, W.; Aebersold, R. Chemical probes and tandem mass spectrometry: a strategy for the quantitative analysis of proteomes and subproteomes. Curr. Opin. Chem. Biol. 2004, 8, 66-75. (6) Kenyon, G. L.; Bruice, T. W. Novel sulfhydryl reagents. Methods Enzymol. 1977, 47, 407-430. (7) Gygi, S. P.; Rist, B.; Gerber, S. A.; Turecek, F.; Gelb, M. H.; Aebersold, R. Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nat. Biotechnol. 1999, 17, 994999. (8) Sechi, S.; Chait, B. T. Modification of cysteine residues by alkylation. A tool in peptide mapping and protein identification. Anal. Chem. 1998, 70, 5150-5158. (9) Sechi, S. A method to identify and simultaneously determine the relative quantities of proteins isolated by gel electrophoresis. Rapid Commun. Mass. Spectrom. 2002, 16, 1416-1424. (10) Shen, M.; Guo, L.; Wallace, A.; Fitzner, J.; Eisenman, J.; Jacobson, E.; Johnson, R. S.; Isolation and isotope labeling of cysteine- and methionine-containing tryptic peptides: application to the study of cell surface proteolysis. Mol. Cell. Proteomics 2003, 2, 315324. (11) Patterson, S. D. From electrophoretically separated protein to identification: strategies for sequence and mass analysis. Anal. Biochem. 1994, 221, 1-15. (12) Rauch, A.; Bellew, M.; Eng, J.; Fitzgibbon, M.; Holzman, T.; Hussey, P.; Igra, M.; Maclean, B.; Lin, C. W.; Detter, A.; Fang, R.; Faca, V.; Gafken, P.; Zhang, H.; Whiteaker, J.; States, D.; Hanash, S.; Paulovich, A.; McIntosh, M. W. Computational Proteomics Analysis System (CPAS): an extensible, open-source analytic system for evaluating and publishing proteomic data and high throughput biological experiments. J. Proteome Res. 2006, 5, 112121. (13) Nesvizhskii, A. I.; Keller, A.; Kolker, E.; Aebersold, R. A statistical model for identifying proteins by tandem mass spectrometry. Anal. Chem. 2003, 75, 4646-4658. (14) Haab, B. B.; Geierstanger, B. H.; Michailidis, G.; Vitzthum, F.; Forrester, S.; Okon, R.; Saviranta, P.; Brinker, A.; Sorette, M.; Perlee, L.; Suresh, S.; Drwal, G.; Adkins, J. N.; Omenn, G. S. Immunoassay and antibody microarray analysis of the HUPO Plasma Proteome Project reference specimens: systematic variation between sample types and calibration of mass spectrometry data. Proteomics 2005, 5, 3278-3291. (15) Bellew, M.; Coram, M.; Fitzgibbon, M.; Igra, M.; Randolph, T.; Wang, P.; May, D.; Eng, J.; Fang, R.; Lin, C.; Chen, J.; Goodlett, D.; Whiteaker, J.; Paulovich, A.; McIntosh, M. A suite of algorithms for the comprehensive analysis of complex protein mixtures using high-resolution LC-MS. Bioinformatics 2006, in press. (16) Molloy, M. P.; Donohoe, S.; Brzezinski, E. E.; Kilby, G. W.; Stevenson, T. I.; Baker, J. D.; Goodlett, D. R.; Gage, D. A. Largescale evaluation of quantitative reproducibility and proteome coverage using acid cleavable isotope coded affinity tag mass spectrometry for proteomic profiling. Proteomics 2005, 5, 12041208. (17) Ong, S. E.; Blagoev, B.; Kratchmarova, I.; Kristensen, D. B.; Steen, H.; Pandey, A.; Mann, M. Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol. Cell. Proteomics 2002, 1, 376-386. (18) Arnott, D.; Kishiyama, A.; Luis, E. A.; Ludlum, S. G.; Marsters, J. C., Jr.; Stults, J. T. Selective detection of membrane proteins without antibodies: a mass spectrometric version of the Western blot. Mol. Cell. Proteomics 2002, 1, 148-156. (19) Zhang, R.; Sioma, C. S.; Thompson, R. A.; Xiong, L.; Regnier, F. E. Controlling deuterium isotope effects in comparative proteomics. Anal. Chem. 2002, 74, 3662-3669.

PR060102+

Quantitative Analysis of Acrylamide Labeled Serum Proteins by LC

Recommend Documents